The examples he presents of false text can all basically be discriminated from human generated text by looking for logical consistency. I wonder if this will be something attention focuses to within schools. With all these fake news articles, spam emails, and fake viral images I think this new generation should be trained to recognize these things early on. Even if we can’t make discriminating classifiers through technology as of yet, we can hopefully train ourselves to discriminate between fake and real media in the meantime.